Python UNICODE GBK UTF-8 之間相互轉換

import urllib

rawdata = urllib.urlopen('http://www.google.cn/').read()

import chardet

print chardet.detect(rawdata)

輸出結果是：

    {'confidence': 0.98999999999999999, 'encoding': 'GB2312'}

# win下命令列引數為gbk編碼：star.gbk2unicode(sys.argv[1]) + u'也有'

def gbk2unicode(s):

    return s.decode('gbk', 'ignore')

# 指令碼檔案#coding:utf-8時預設不帶u的字串為utf8字串：star.utf82unicode('我')

def utf82unicode(s):

    return s.decode('utf-8', 'ignore')

# 帶u的字串為unicode

# star.unicode2gbk(u'\u4e5f\u6709')

# star.unicode2gbk(u'也有')

def unicode2gbk(s):

    return s.encode('gbk')

# 帶u的字串為unicode

# star.unicode2utf8(u'\u4e5f\u6709')

# star.unicode2utf8(u'也有')

def unicode2utf8(s):

    return s.encode('utf-8')

# win下命令列引數為gbk編碼：star.gbk2utf8(sys.argv[1]) + '也有'

def gbk2utf8(s):

    return s.decode('gbk', 'ignore').encode('utf-8')

def utf82gbk(s):

    return s.decode('utf-8', 'ignore').encode('gbk')